Goals:
- Create training data for text recognition tasks
- Use custom / video game fonts
- Minimal human intervention for maximum generation
docker-compose up
Runs as root as to add font files.
Volume mount provides out
directory of images for you.
Use docker-compose up --build
if dependencies change.
Watchdog might not work on some systems. Reloading the webpage manually should work.